The Learning Corner by Precursor
Welcome to the Learning Corner, a weekly Precursor Ventures podcast, where members of the Precursor team walk through their favorite articles and news snippets across the venture ecosystem.
Episode #54: Teaching Students to Challenge AI, Is AI The New Shadow Bank?
First up, we dive into “Beyond True or False: Teaching Students to Interrogate AI Unreliability”, a Substack by Nick Potkalitsky, which proposes a new framework—borrowed from literary theory—for teaching students to critically evaluate AI-generated content. We discuss how this lens can help people move beyond simple trust/distrust binaries and become better co-creators with AI. Then, we explore “Is AI the New Shadow Bank? (Yes…)”, a piece that draws parallels between today’s AI economy and the pre-2008 shadow banking system. Instead of mortgage-backed securities, today’s collateral is GPU access, compute contracts, and foundation models—and we ask: is the real innovation the credit system AI is built on?
Key Points
- Applying a critical lens to AI outputs, similar to literary critique, can help users better understand and evaluate the reliability of AI-generated content.
- The circular nature of AI infrastructure investments and their potential financial instability may parallel the risks seen in the shadow banking sector before the 2008 financial crisis.
- Building effective AI tools requires explicit, detailed instructions, and this process can highlight and refine underlying assumptions and methods.
Chapters
| 0:00 | |
| 0:30 | |
| 4:50 | |
| 9:19 | |
| 15:13 |
Transcript
Loading transcript...
- / -